Autism has a devastating effect on language development. Fifty percent of persons with autism do not develop complex, communicative speech, and lack of speech is associated with very poor outcomes in autism. Yet there is evidence that specific treatments delivered early in life can increase the number of children with autism who can learn useful speech. Various early intervention projects have indicated that 75-95 percent of children receiving very intensive and carefully constructed early intervention programs develop some useful speech by age 5. However, we have very few manualized, empirically demonstrated treatments to guide practitioners in developing expressive speech in young children with autism. The mechanisms that underlie the language impairment in autism are not clear. Lack of social engagement and mental retardation are assumed to play pivotal roles and treatments have generally focused on increasing social engagement, imitation skills, and understanding of language in order to develop language. The approach used to teach speech and language in the Denver Model is built from such a conceptualization. However, in the past few years, empirical findings have begun to suggest that the presence of oral-motor dyspraxia in at least some children with autism may be an additional barrier to speech development. A promising treatment for oral motor dyspraxia solidly based in neuromotor aspects of speech production has been successfully used for the past 20 years: the PROMPT approach. While originally developed for other neurologically based speech impairments, it has begun to be used clinically successfully with young nonverbal children with autism after more traditional approaches to speech development have failed. The main objectives of the project are to complete standardization and refinement of these two treatments for expressive language in autism: PROMPT and the Denver Model, and to examine their effects on nonverbal preschoolers with autism by obtaining initial pilot data on efficacy of each for developing expressive speech. These activities are prerequisite for a full-scale experimental treatment study comparing these two treatments for developing speech in young nonverbal children with autism. (1) Treatment manuals both for therapists and for parents will be completed, tested, and refined. (2) Treatment fidelity measures will be developed and tested for both therapy and home components. (3) Pilot testing will be used to establish appropriate "dosages" of treatments to assure reasonable progress. (4) Finally, a short-term pilot study using random assignment of matched participants and single subject design will be conducted to examine efficacy of each approach.